This PDF file contains the editorial “A Look Back” for OE Vol. 48 Issue 10

We experimentally demonstrate the fabrication of silicon optical fibers by using the powder-in-tube technique. The fibers are drawn from a preform utilizing a custom-made fiber drawing system. Silicon optical fibers having cladding diameters in the range of 40 to 240 μm, core diameters in the range of 10 to 100 μm, and an approximate overall length of 7 cm have been fabricated. The powder-in-tube technique is versatile and can be utilized to fabricate fibers with different dimensions and core/cladding materials.

We propose and demonstrate a novel approach to identify linear and nonlinear propagation regimes of an optical signal in an optical fiber link by using chaos analysis. We show that the chaotic characteristics of a propagating optical signal are affected by both the chromatic dispersion and the nonlinear effects in the optical fiber. Linear or nonlinear behavior is detected by determining the maximum Lyapunov exponent of the signal and the use of the recurrence plot technique. An experimental demonstration is performed using 10-Gbps signal propagation in a 100-km fiber link with different launched optical powers. Chaos analysis shows a clear identification of the linear and nonlinear optical propagation regimes by using a classification scheme based on a multilayer neural network. Numerical simulations confirm the experimental results.

A fiber-optic bundle-based submillimeter probe for endoscopic common-path optical coherence tomography with focusing system is presented. Three different focusing systems, bulk lenses, a gradient index (GRIN) lens, and a microlens array, are proposed and compared. The simulation results show that all the focusing systems can focus the beam with different degrees of effectiveness. The major differences between the different focusing systems exist in the beam size and the divergence angle between the inner and outer pixels of the fiber bundle. The microlens array shows the smallest change in beam size from the central fiber to the outermost fiber, while the GRIN lens exhibits the largest. Signal-to-noise ratio of the system, taking attenuation and cross talk between adjacent pixels of the fiber bundle into account, is also analyzed.

The NASA Mars Science Laboratory (MSL) rover is the third-generation robotic vehicle planned for Mars exploration, scheduled to land on Mars in 2012. There are 15 cameras on board the rover, 10 of which have engineering functions and 5 of which are designed for scientific investigations. Four of the five science cameras are built by Malin Space Science Systems to acquire color images via an RGB Bayer pattern 1600×1200 pixel charge-coupled device. These cameras are categorized into three types: Mast Camera (Mastcam), of which two different versions will be flown; Mars Hand Lens Imager (MAHLI); and Mars Descent Imager (MARDI). These cameras will investigate Martian geology and geomorphology and provide the first color views of the Martian surface that are identical to the natural color provided by typical consumer digital cameras. The author designed and built the lenses for these MSL science cameras, Mastcam utilizes a 34-mm focal length F/8 and a 100-mm focal length F/10 lens; MAHLI utilizes a lens that varies from a 18.4-mm focal length F/9.8 (close focus) to 21.4-mm focal length F/8.5 (infinity focus); and MARDI utilizes a 9.7-mm focal length F/3 lens. The requirements, design, and performance of these lenses are described.

Wear profiles obtained with tools that oscillate in one direction and simultaneously displace perpendicularly in another are classified and analyzed. Profiles are obtained for different-shape tools and oscillation amplitudes. Analytical and numerical calculations of wear were made with known rectangular, elliptical, and diamond tool shapes. In the test, both sets of calculations turned out to be equal. A generalized equation was used to describe and speed up numerical calculations of the inner points of both known and other asymmetrical tools. The wear profiles obtained are of great help in enabling us to determine the shape, size, and oscillation amplitude of a tool to achieve a desired wear. Experimental results show that these simulated profiles can be obtained in practice using square, circular, and diamond tool shapes.

A dual-modality fiber scope probe that can simultaneously obtain the spectroscopic signature and images is proposed and illustrated. The photonic crystal fiber and micro CCD camera form the probe distal end. A preliminary experimental investigation using the developed probe has been carried out on phantom tissue stained with fluorophores. Both hollow-core and double-clad photonic crystal fibers have been used at the distal end in separate embodiments, and their collection efficiencies are compared. This proposed probe and methodology are expected to find potential biomedical diagnostics applications.

Shack-Hartmann wavefront sensing is the most widely used technique for the measurement of wavefront distortion in adaptive optics (AO). Assembly errors decrease the accuracy of the Shack-Hartmann wavefront sensor and degrade the quality of the AO system. The slant correction method based on principal component analysis (PCA), which has its origin in the digital graphic discipline, is introduced in an AO system. The slant correction method based on PCA and the results of this method for adaptive optics system are investigated experimentally. The experiment results show that the AO system becomes less sensitive to noise, and no high assembly precision is required with the slant correction based on PCA.

A method based on basic phase measuring deflectometry is proposed for testing the aspherical mirror. The method uses a reference screen in two different distances from the mirror under test. The sinusoidal, intensity-modulated patterns generated by the computer are displayed on the LCD screen, and the camera observes the patterns reflected off the testing mirror. The observed pattern appears distorted depending on the shape of the mirror. Using the phase-shifting technique, the original ray of every image point and its corresponding deflected ray can be constructed. Their intersection points and the surface normal are obtained. Then the mirror surface is reconstructed with high accuracy by numerically integrating the surface normals. The proposed method is robust against noise and can test the mirror full field. In this work, the method is introduced, and computer simulation and experimental results are shown.

Laser-ablation based spectrochemical analytical methods could be used in applications in which the capability for spatially resolved analysis is required. Such analysis gives a 2-D or 3-D map of the monitored chemical elements. Here we introduce automation of a 2-D surface analysis in the laser-induced breakdown spectroscopy setup by implementation of an autofocus algorithm based on the evaluation of the image sharpness. The most suitable algorithm with respect to its speed, accuracy, and durability against digital noise is chosen by testing different gradient-based methods and methods working in the frequency domain.

We present a new single-shot 3-D shape measurement method using a digital micromirror device (DMD) camera and fringe projection. In optics of the DMD camera, individual DMD mirrors operate as controllable high-speed shutters for corresponding charge-coupled device (CCD) pixels. Therefore, four phase-shifted images can be recorded within one frame of the CCD camera by integrated phase-shifting methods using correlations. To obtain 3-D information of the object, an easy and accurate phase-to-3-D calibration method is performed. The principle and hardware are presented. The experimental results under dynamic conditions show that 3-D shape information can be analyzed from only a single image.

An emission layer with the poly-vinylcarbazole (PVK) and Ir(ppy)3[tris(2-phenylpyridine)iridium(III)] as the host and guest materials, respectively, was prepared on a PEDOT:PSS/ITO substrate. TPBI [2,2′,2″-(1,3,5-benzinetriyl)-tris(1-phenyl-1-H-benzimidazole)] was introduced to improve the light efficiency of the devices. Green phosphorescence polymer LEDs with the structure ITO/PEDOT:PSS/PVK:Ir(ppy)₃/TPBI/LiF/Al were fabricated and their electrical and optical properties investigated at various doping concentrations (0.5 wt % to 2.5 wt %) of Ir(ppy)₃. The maximum luminance and current density were about 8600 cd/m² and 155 mA/cm² at 7 V, respectively, when the TPBI layer was introduced between the emission layer and cathode electrode film. The emission spectrum and Commission International d'Eclairage (CIE) color coordinates were about 512 nm and (x,y=0.27 to 0.28, 0.61 to 0.63), respectively, showing a green color.

We propose a molecular dynamics method dot generation scheme incorporating the periodic boundary and fixed-cell boundary geometrical conditions to generate random dot distribution on the order of several million dots for use in the base of a thin light guide. In addition, the generation scheme can be used to easily adjust the dot density at any local region from illuminance causing feedback in the optical design phase. Based on the incorporation of these two geometrical conditions, the generation of the random dots for the whole LED light guide can be simplified to just three subdomains, one domain represented by the periodic boundary condition, and the other two domains represented by the fixed-cell boundary condition. Employing these conditions significantly shortens the computational time necessary to obtain optimal random dot distribution and subsequent optical optimization on a PC using optical commercial software.

Switchable single-longitudinal-mode (SLM) dual-wavelength erbium-doped fiber laser at room temperature is demonstrated. One fiber Bragg grating (FBG) directly written in a polarization-maintaining and photosensitive erbium-doped fiber as the wavelength-selective component is used in a linear laser cavity. Because of the polarization hole burning enhanced by the polarization-maintaining FBG, the laser can be designed to operate in stable dual-wavelength or wavelength-switching modes with a wavelength spacing of 0.202 nm by adjusting a polarization controller. The stable SLM operation is guaranteed by a saturable absorber. The optical signal-to-noise ratio of the laser is >40 dB. The amplitude variation in nearly 1.5 h is <0.5 dB for both wavelengths.

The output beam characteristics of a toric concave mirror laser resonator are discussed. The experiments of the toric concave mirror laser resonator, the plane-concave stable resonator, and the parallel plane resonator with the same Fresnel number of 17.25 are done on HUST2000 high-power transverse flow CO₂ laser. The output laser beam of 1820 W with a doughnutlike distribution, the output laser beam of 1860 W with a Laguerre-Gaussian (LG) transverse electromagnetic modes (TEM)₆₀ distribution and the laser beam of 1000 W with are approximate LG TEM₃₀ distribution are obtained, respectively, at the same discharge current of 5 A. The scanning welding test on 316L stainless steel shows that the output beam of the toric concave mirror laser resonator can used for deep penetration laser welding, the welding depth is >2.4 mm at a velocity of 1 m/min, whereas that of other two resonators cannot be used for deep-penetration laser welding because of low beam quality or low output power. This resonator can be used in the laser with a large transverse gain size to improve beam quality greatly without sacrificing the output power.

A dye ring laser is stabilized to a D_2a Doppler-free feature of sodium vapor using a LabVIEW^®-based, phase-sensitive servo. Locking precision and stability, at better than ±1 MHz, are suitable for Na lidar applications. This performance was achieved with improved digital filtering and new approaches to the problem. The inverse (type II) Chebyshev discrete filter employed demonstrates superior filtering and computational efficiency plus improved flexibility. New approaches include the determination of optimum modulation frequency, laser-tuning sensitivity, and bandwidth requirements via spectral analyses of the noise spectrum, derivative scan, and modulated spectrum. This practice guides a user in selecting the system operation parameters and negotiating the trade-offs involved when expanding the filter's passband. Allan deviation plots provide a quantitative description of the short- and long-term frequency excursions. A comparison of Allan deviation plots before and after locking shows a substantial improvement in stability throughout time scales from 0.10 to 10 s.

A method for generation of a time- and wavelength-interleaved pulse train is demonstrated, that can be used to attain a multiwavelength pulse train with a 40-Gbps or a potentially even higher repetition rate. This method is highly flexible because the repetition rate, the intensity, and pulse width of each wavelength and the time interval between adjacent wavelengths can be readily and independently adjusted. A time- and wavelength-interleaved pulse train with a repetition rate of 40-Gbps is experimentally demonstrated. Potential of generating a multiwavelength pulse train with a 100-Gbps repetition rate is also discussed.

Nd³⁺ :Er³⁺:Yb³⁺ codoped Al₂O₃ powder is prepared by the nonaqueous sol-gel method and pressed into a small wafer. The two-stage sensitization processes for energy transfer among Yb³⁺, Er³⁺, and Nd³⁺ are discussed. Under a 978-nm semiconductor laser excitation, the photoluminescence intensities at 813 and 887 nm, corresponding to the ⁴F_5/2+²H_9/2 →⁴I_9/2 and ⁴F_3/2 →⁴I_9/2 transitions of Nd³⁺ ions, enhance monotonously with the increasing temperature, and the fluorescence intensity ratio of the near-IR emissions can be fitted the exponential function R=3.12 exp(-1150/T) in the temperature range of 375 to 889 K. The maximum sensitivity is 0.0015 K^-1 at 574.3 K. The data are of undoubted repeatability, and the material is applicable for sensitive, high-temperature measurement.

A fast, sensitive temperature sensor using changes in extraordinary optical transmission (EOT) through an array of nanoholes in a metallic film with a monochromatic collimated light source is analyzed and experimentally demonstrated. The device is composed of a dielectric material in contact with a metallic film with an array of nanoholes, (25 arrays per chip were treated as a single sensor), on a glass substrate. Although the present experiment used a volume of water in a test cell to demonstrate the phenomena, there are many configurations that implement this technology in a variety of applications. The fast response of the device is expected because it measures the average temperature of a thin layer of the dielectric material directly above the nanoholes, ~100 nm. The thermal response parameter of the nanohole array sensing region is ~10⁸ times smaller and orders of magnitude faster than that of the reference thermistor. The increased speed and sensitivity of the nanohole array temperature sensor over the thermistor is observed in the reported data. The configuration of this temperature sensor was used for simplicity and to demonstrate that EOT effects can be used for temperature sensing.

We analyze a highly compact optical add-drop filter topology based on a pair of microdisk resonators and a bus waveguide intersection. The filter is further assessed on an integrated optical 4×4 network for optical on-chip communication. The proposed network structure, as compact as 50×50 μm, is fabricated in a CMOS-compatible process on a silicon-on-insulator (SOI) substrate. Finally, the experimental results demonstrate the proper operation of the fabricated devices.

A receiver configuration using a single shared delay interferometer (DI) for optical differential phase-shift keying (DPSK) wavelength-division multiplexing (WDM) signals is demonstrated and investigated in 40-Gbit/s signal transmission. Experiment comparison between the proposed and conventional configurations is performed for signal performance and a 3-dB bandwidth of the receiver sensitivity penalty. The wavelength mismatch of optical components used and the effects are measured to investigate the wavelength dependency of the proposed configuration. This demonstrates that there is no difference in performance between the two configurations and between WDM channels in the proposed configuration. Thus, it is suggested that the proposed configuration using a single DI can be applied for 20 WDM channels between 1531.116 and 1561.419 nm, with 200-GHz channel spacing in a bandwidth of >30 nm, for receiving DPSK signals.

We propose and demonstrate a novel approach to identify linear and nonlinear propagation regimes of an optical signal in an optical fiber link by using chaos analysis. We show that the chaotic characteristics of a propagating optical signal are affected by both the chromatic dispersion and the nonlinear effects in the optical fiber. Linear or nonlinear behavior is detected by determining the maximum Lyapunov exponent of the signal and the use of the recurrence plot technique. An experimental demonstration is performed using 10-Gbps signal propagation in a 100-km fiber link with different launched optical powers. Chaos analysis shows a clear identification of the linear and nonlinear optical propagation regimes by using a classification scheme based on a multilayer neural network. Numerical simulations confirm the experimental results.

We investigate, design, and develop a prototype real-time synchronous receiver for the second-generation quantum communicator recently developed at the National Aeronautics and Space Administration (NASA) Glenn Research Center. This communication system exploits the temporal coincidences between simultaneously fired low-power laser sources to communicate at power levels several orders of magnitude less than what is currently achievable through classical means, with the ultimate goal of creating ultra-low-power microsize optical communications and sensing devices. The proposed receiver uses a unique adaptation of the early-late gate method for symbol synchronization and a newly identified 31-bit synchronization word for frame synchronization. This receiver, implemented in a field-programmable gate array (FPGA), also provides a number of significant additional features over the existing non-real-time experimental receiver, such as real-time bit error rate (BER) statistics collection and display, and recovery and display of embedded textual information. It also exhibits an indefinite run time and statistics collection.

Two schemes of all-optical tunable delay line are proposed and experimentally demonstrated based on wavelength conversion in reflective semiconductor optical amplifiers (RSOAs) and group-velocity dispersion (GVD) in optical fiber. In the first scheme, large optical delays are shown for a 10 Gbits/s non-return-to-zero (NRZ) signal at short wavelength in the C band. In the second scheme, large optical delays are obtained in the whole C band. The proposed schemes achieve continuously tunable delays with nearly no pulse broadening and very little spectral distortion.

We present a modeling of the Raman amplification of various types of fibers employed for transmission and dispersion compensation in cascaded optical amplified transmission systems including (1) the physical properties and operational characteristics and (2) its implementation in the lightwave-modulated envelop propagation equation for transmission of 40 Gbits/s signals in mutliplexed wavelength optical systems. We demonstrate, by simulation using MATLAB Simulink, the integration of the Raman amplification gain and noise into the transmission system by (1) numerical techniques to solve these design parameters for a single channel bidirectionally pumped system operating in the C-band (1525 to 1565 nm) range at 1550 nm; (2) evaluations of the amplifier parameters such as the gain, the noise figure, and the dispersion factor of distributed and discrete Raman configurations. We design an optimum condition for hut-skipped undersea long-haul transmission using a hybrid combination of Raman and Er-doped fiber amplifers. The long-haul numerical solution is adapted from the average power to a Gaussian pulse operating at 40 Gbits/s and the propagation distributed along the fiber for integration into differential phase shift keying modulation format optical signals using models based on MATLAB Simulink.

A novel optical method for ultrawideband (UWB) monocycle pulse generation and bipolar coding is proposed and experimentally demonstrated. The proposed scheme is based on the cross-polarization modulation effect in highly nonlinear photonic crystal fiber (HNPCF). The monocycle pulse with a time duration of 84 ps is generated, and four groups of bipolar codes are obtained.

Phase noise characteristics of the carrier transmitted by the double-sideband-carrier-suppressed (DSB-CS) modulation system are investigated. On the basis of the established DSB-CS modulation system, the phase noise of the input carrier and the output doubled frequency carrier is measured. The influencing factors on the far-from-carrier phase noise are discussed in detail. The noise current mean-square value of the signal-spontaneous beat noise is first derived in the case that there are two first-order optical sidebands generated by DSB-CS modulation. It is shown that the far-from-carrier phase noise is degraded primarily due to the signal-spontaneous beat noise and the relative-intensity noise.

An approach to improve the coupling efficiency of bidirectional optical subassembly (BOSA) modules is proposed and experimentally demonstrated. We analyzed the wavefront aberration coefficients of a typical BOSA. It was found that the 45-deg wavelength filter induces coma and astigmatism, and then it further deteriorates the laser diode to fiber coupling. We measured the BOSA efficiencies based on a series of different filters. For a typical 0.5-mm filter, 25% coupling efficiency improvement was achieved by optimizing the filter parameters.

We demonstrate a modulated retroreflector that utilizes large-area multiple quantum well modulators on all three faces of a retroreflector. The large-area devices, fabricated by metalorganic chemical vapor deposition, are characterized in terms of the yield and leakage currents. A yield higher than that achieved previously using devices fabricated by molecular beam epitaxy is observed. The retroreflector module is constructed using a standard FR4 printed circuit board (PCB) technology, thereby simplifying the wiring issue. A high optical contrast ratio of 8.23 dB is observed for a drive of 20 V. A free-standing PCB retroreflector is explored and found to have insufficient angular tolerances (±0.5 deg). We show that the angular errors in the corner-cube construction can be corrected for using off-the-shelf optical components as opposed to mounting the PCBs on a precision corner cube, as has been done previously.

We address a framework to reduce the unwrapping errors of the measurement system using a digital multiple-wavelength phase-shifting algorithm. In particular, the following framework is proposed: (1) smooth the raw phase by smoothing the sine and cosine images of the phase computation of the inverse tangent function; (2) locate and remove the incorrectly unwrapped points by the monotonicity condition of the phase map; (3) obtain the unwrapped phase map for the shortest wavelength without smoothing; (4) detect holes and fill them to preserve as much useful information as possible. Experiments demonstrated that the proposed framework significantly alleviated the measurement errors caused by the phase noise.

We describe the design and implementation of a long-wave infrared (LWIR) spectrometer based on the spatial heterodyne spectroscopy (SHS) technique, and present the first measurements of broadband LWIR spectra taken with an SHS instrument. This work represents the first successful application of SHS to the field of LWIR spectroscopy, which is currently dominated by Fourier transform spectrometers, grating spectrometers, and Fabry-Perot interferometers. A unique combination of properties makes SHS a valuable addition to the existing types of LWIR spectrometers. Most notable are the interferometric throughput (no slit), lack of moving parts, and that the measured spectra are not contaminated by a changing scene, which makes SHS particularly suitable for applications in rugged environments and on moving platforms. The instrument discussed here is called the Spatial Heterodyne Imager for Chemicals and Atmospheric Detection (SHIMCAD), and is designed to cover the wavelength range between about 8.4 μm (1190 cm^-1) and 11.2 μm (890 cm^-1) with a spectral resolution of about 4 cm^-1. First, laboratory SHS transmission spectra of methanol and polyimide (Kapton^®) are presented. The instrument is built to be mobile, so that ultimately field measurements can be conducted.

The tomographic imaging technique in digital holography is improved and applied to the measurements of the refractive index distributions of two kinds of optical fiber. They are reconstructed from a set of quantitative phase maps obtained with a digital holographic method as the object is rotated for 180 deg. The errors of the phase maps caused by the defocusing due to the unwanted shift of the object relative to the rotation axis are corrected by numerical focus adjustment. The tomographic images of the fiber samples show the effectiveness of our method.

The expressions for the reconstructed field from the sample of the diffracted wave, which is produced by illuminating an object, are found by use of different diffraction integrals in the digital holography. The numerical reconstruction methods that truncate and sample this field are compared in overlapping quality, accuracy, pixel resolution, computation window, and speed. The fast Fourier transform (FFT)-based direct integration method for the Fresnel integral and the modified FFT-based direct integration method for the Rayleigh-Sommerfeld integral have similar overlapping quality and can flexibly control pixel resolution and computation window size. Meanwhile, the FFT-based angular spectrum method is superior to the FFT-based convolution method in accuracy and speed. The experimental results are presented to verify these consequences.

Bone age assessment of children is a procedure frequently performed in pediatric radiology. The feature extractions of metaphyseal and epiphyseal regions are crucial to automatic bone age assessment. The first step of feature extraction is applying a segmentation scheme to find exact regions of epiphysis and metaphysis. A segmentation method is normally based on both intensity information and the relative location of pixels. There is a fundamental problem; when the intensity contrast of soft tissue and bony tissue is poor, bony and soft tissue cannot easily be separated. We propose an α-gamma equalization method to increase the intensity contrast between bony and soft tissue. Sobel, two-means, Canny edge-detection, and watershed methods are applied to illustrate the effect of this method on initial segmentation. Adaptive two-means and gradient vector flow snake are adopted for the final segmentation. Experimental results reveal that α-gamma equalization-enhanced two-means initial segmentation with an adaptive two-means clustering scheme can greatly reduce the average error measurements of segmentations. These are evaluated according to the following measurements: misclassification error, edge mismatch, region nonuniformity, relative foreground area eror, and modified Hausdorff distance. Furthermore, the experimental results show that the proposed scheme provides increased stable performance for the segmentation of epiphyseal/metaphyseal regions.

We present an adaptive tone reproduction algorithm for displaying high-dynamic-range (HDR) images on conventional low-dynamic-range (LDR) display devices. The proposed algorithm consists of an adaptive tone reproduction operator and chromatic adaptation. The algorithm for dynamic range reduction relies on suitable tone reproduction functions that depend on histogram-based parameter estimation to adjust luminance according to global and local features. Instead of relying only on reduction of dynamic range, this chromatic adaption technique also preserves chromatic appearance and color consistency across scene and display environments. Our experimental results demonstrate that the proposed algorithm achieves good subjective quality while preserving image details. Furthermore, the proposed algorithm is simple and practical for implementation.

We present a novel objective full-reference image quality assessment metric based on multiscale geometric analysis. The multichannel behavior of the human vision system is emulated by contourlet transform, a perceptual subband decomposition. Not only the contrast-masking effect but also the entropy-masking effect is considered to deal with the visual masking issue. In the error pooling stage, the frequency sensitivity of the HVS is investigated. Nonlinear and linear fusion schemes of subband distortion are compared. Extensive validation experiments are performed on two professional image databases, the LIVE database supplied by the University of Texas and the A57 database supplied by Cornell University. Compared with several state-of-the-art image quality metrics, the proposed metric demonstrates improvement on prediction accuracy and robustness.

We present a detailed experimental study of circle-marker detection algorithms based on color information. Three color spaces-HSV (hue, saturation, value), normalized color space rgb, and the gray-world normalization (NG) color space-are of main concern. We compare the algorithms based on these color spaces and combine them to use the color information simultaneously with shape-based marker detection techniques, and the Hough transformation, to achieve a new, robust, color-based circle-marker detection algorithm that gives higher accuracy. Experimental results show that the proposed algorithm is good enough to detect a circle marker in a complex background with various illuminations.

Results obtained by current foreground segmentation algorithms are often corrupted with splits and defects, making them undesirable for high level applications. We present a new segmentation algorithm that takes advantage of tracking information for segmentation. First, the moving object is tracked with a mean-shift tracker. Second, the tracking information is used to construct a tracking-based appearance model, which is then used for segmentation. Finally, a mixture model is used to cope with the tracking error. Experiments show that the proposed method provides improved results in changing settings.

The increasing prevalence of obesity suggests a need to develop a convenient, reliable, and economical tool for assessment of this condition. Three-dimensional (3-D) body surface imaging has emerged as an exciting technology for the estimation of body composition. We present a new 3-D body imaging system, which is designed for enhanced portability, affordability, and functionality. In this system, stereo vision technology is used to satisfy the requirement for a simple hardware setup and fast image acquisition. The portability of the system is created via a two-stand configuration, and the accuracy of body volume measurements is improved by customizing stereo matching and surface reconstruction algorithms that target specific problems in 3-D body imaging. Body measurement functions dedicated to body composition assessment also are developed. The overall performance of the system is evaluated in human subjects by comparison to other conventional anthropometric methods, as well as air displacement plethysmography, for body fat assessment.

We propose a novel large-scale conditional random field model with respect to the problem of natural outdoor scene labeling. The novelty of the proposed method lies in three aspects: 1. features from two neighboring regions are concatenated to form the input of the pair-wise classifier to compensate for the simultaneous feature deviation of neighboring regions; 2. the definition of a generalized neighboring system and the incorporation of direction-specific patterns in conditional random field models based on the generalized neighboring system to better simulate the visual cognition of human being; and 3. the definition of a similarity criterion based on the bags-of-words expression to facilitate the incorporation of semantic patterns. The proposed model is first evaluated over the Corel dataset. Both qualitative and quantitative results show that our model is capable of modeling large-scale spatial relationships between objects in natural outdoor scenes, and achieves better results than other existing conditional random field models. Furthermore, our model is also evaluated over several other natural datasets, which are taken from logged field tests, to further demonstrate the adaptability of our model to different lighting conditions.